Simon Blomberg <s.blomberg1 <at> uq.edu.au> writes: > > To get a confidence interval on lambda, you need to have measures of variability in the elements of the > transition matrix. If you have that, you can use a parametric bootstrap to get approximate confidence > intervals. I have done this, and it seems to work. Alternatively, you could calculate a Bayesian > posterior density for lambda using the Bayesian melding methods developed by Adrian Raftery et al., and > calculate an HPD interval from that. I've done that too. It's slightly more difficult, however. > > Simon.
Or use the delta method: Skalski, John R., Joshua J. Millspaugh, Peter Dillingham, and Rebecca A. Buchanan. 2007. Calculating the variance of the finite rate of population change from a matrix model in Mathematica. Environmental Modelling & Software 22, no. 3 (March): 359-364. http://www.sciencedirect.com/science/article/B6VHC-4JMM5XY-1/2/a698149bc3798c273766cfacdf40bba5 (accessed August 29, 2007). I've written a little bit of generic delta-method code, but I don't know if it's this generic. Ben Bolker ______________________________________________ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.